Due to recent development of digital image acquisition and processing technologies, a variety of image photographing has been made through the use of a digital camera and a camcorder. In other words, still image or moving image for a short time is photographed using a digital camera, while various moving images are photographed using a camcorder.
Also, development has been made from simple two-dimensional (2D) image photographing to 3D image photographing owing to evolution of image photographing technique. Namely, the photographing technique has been developed from 2D image photographing that acquires image information on a plane using a single fixed camera to 3D image photographing that acquires multi-viewpoint images while moving camera.
A typical example of 3D multi-viewpoint image acquisition techniques is panoramic image photographing technique. A part of existing digital cameras offers a panoramic image photographing function. This panoramic image photographing function allows a user to photograph a plurality of images by moving little by little a camera horizontally or a photographing angle, and then to obtain one large image by stitching the photographed images.
3D image acquisition methods are largely classified into three types: parallel type, convergent type and divergent type. FIG. 1 shows each of the above three types.
Among the image acquisition methods shown in FIG. 1, a panoramic image acquisition method is limited to a case of photographing in the parallel type or divergent type. A principal goal to make the panoramic image is to compose images for ambient environment and scenery at various viewpoints.
The convergent type photographing technique in the 3D multi-viewpoint image acquisition is a technique which concentrates photographing viewpoints to a center, which is opposite of the divergent type and photographs an object while moving camera in circular manner. By this means, various viewpoints of an object (e.g. commercial goods for the internet shopping) can be provided to users by photographed images, and users can control their viewpoints as they want. FIG. 2 shows an example that photographs an object in the convergent type.
Meanwhile, to acquire a high quality 3D multi-viewpoint image, trembling of hands and movement of camera should be minimized on photographing. In the process of generating a panoramic image in the parallel and divergent type photographing ways, compensation for tremble of camera is conducted in a stitching algorithm after photographing. Even with this compensation process, however, stable photographing should be premised to obtain the high quality of panoramic image.
Further, in case of acquiring 3D multi-viewpoint images in the convergent type, it is not easy to process with respect to tremble and movement which occur in photographing. Since the convergent type multi-viewpoint image is configured in such a way in order for providing different views of an object to a user, there may be a significant inconvenience when the user uses the 3D multi-viewpoint images whose different viewpoint images are not aligned.
For example, in case where the user photographs a specific object while rotating at a predetermined angle only in a horizontal direction where a vertical direction is fixed, the most efficient case may be the case of acquiring image while moving a camera at a predetermined angle in a state where a focus and an zooming of the camera are unchanged. In the case of photographing convergent type multi-viewpoint images, however, acquired images would be photographed in the unstable state due to a tremble of hands and a vertical movement of camera when photographing an object while person moves without using a special device.
To solve the above problem, that is, the tremble phenomenon, images are photographed using a special device, or endeavors have been progressed to minimize the tremble problem of photographed images using a postprocessing technique.
However, equipment with a method for using the special device is costly, and is also employed only in some special companies since there is a spatial limitation for its installation for general users to use. Moreover, a method using a digital image processing techniques by computer has a limitation that the general users cannot easily use it due to the computational complexity and technical limitation.